Making an Aqualung

How to Construct Your Own Underwater

Breathing Apparatus
By E.T. FEARON




UNDERWATER swimming is a sport which is Iast growing in popularity all over the world. In this country its
devotees are limited to about Iour months sea-swimming in each year, but the enchanting new world that is opened
upeven to swimmers oI only average abilitymakes it a sport that is sure oI a large British Iollowing. By wearing
Ilippers and goggles alone, it is possible to view marine liIe clearly and at close quarters, holding the breath to dive in
order to inspect anything which attracts the attention. With the simple addition oI a snorkel tube one caneither lie
motionless on the surIace oI the water or move slowly along, with a gentle movement oI the Ieet, while continuously
watching the colourIul panorama below.
This sport is known as "skin diving"; but any skin diver will tell you that the Iinal thrill oI the sub-aquatic sportsman
comes with the possession oI an aqualung. "Aqualung" is the name given to the selI-contained compressed-air diving
apparatus which permits a trained diver to swim down to two or three hundred Ieet below the surIace, or to explore the
sea-bed in shallow waters Ior 40 minutes at a time. While a new aqualung might cost 40 to 70 it is possible to
assemble at home a perIectly reliable piece oI equipment Ior approximately 14. The work involved is well within the
capability oI the average model engineer, as will be seen Irom the line drawings which illustrate the modiIicationsthat
have to be carried out on easily obtainable parts.
Warning
BeIore going on with the details oI the home-made aqualung it would be as well to point out that, simple as this
equipment is, in the wrong hands, or in inexperienced hands, it could be a source oI serious trouble. A cylinder
explosion could be quite easily Iatal, so there must be no question oI using anything but a new, Iully-certiIicated
cylinder. Air jets, either at 2,000 or 20lb. per sq. in., can also cause bodily harm; they must be treated with the utmost
respect and that simply means that no air must be allowed to escape.
Private pumping to recharge cylinders must not be considered, Ior it is a specialist's job. Empty cylinders should be sent
to a depot oI the British Oxygen Company Ltd., who will ensure that they are tested regularly by hydraulic pressure to
practically twice the working pressure.
BeIore Iinally using the aqualung, take a course oI instruction with your nearest branch oI the British Sub-aqua Club.
The Iee will be money well spent, Ior you will learn not only straightIorward diving, but also what to do should an
emergency arise. AIter graduating to sea diving, always stay as close to the surIace as you can swim with empty lungs,
Ior you may have to do just that one day. And in any case never dive deeper than 60It. nor stay there Ior more than 15
minutes at a time, and you need never Iear the "bends".
The aqualung described here has been thoroughly tested and has been used successIully in many dives; it is believed to
be a saIe and trustworthy piece oI equipment. However, neither the editor, author, nor publisher oI this journal will be
able to accept responsibility Ior any accidental damage to person or property arising Irom the use oI a similar
equipment.



Principle of the Aqualung
The reason there is a limit to the useIul length oI a snorkel tube is that water pressure increases at the rate oI 15lb. per
sq. in. Ior every 33It. oI depth. It is noticeable when "treading water" with the mouth just above the surIace, that
breathing requires a slight eIIort; iI the lungs sink 2It. Iarther below the surIace the pressure oI the surrounding water
prevents them working at all. In order to counteract this eIIect, the aqualung is arranged automatically to adjust the
pressure oI the air it supplies to equal the pressure oI the surrounding water. This is done by the "demand" valve, which
also conserves the air in the cylinder by closing when the lungs are not taking in air.
The cylinder has a capacity oI 0.4cu.It., and it contains 40cu.It. oI air, compressed to a pressure oI 2,000lb. per sq. in.,
when Iully charged. T`his extremely high pressure has Iirst to be reduced by a valve to about 100lb. per sq.in., and the
air is then Ied to the demand valve, which perIorms the dual Iunction oI equalising the pressure in the inlet tube to that
oI the surrounding water and supplying air to the lungs as soon as they start to inhale.
Conventional aqualungs terminate the inlet and exhaust tubes correctly close together, but as the complete valve gear is
mounted in a single unit on top oI the cylinder there is oIten a diIIerence in pressure between the sensitive demand
valve and the diver's mouth. In the normal horizontal swimming position this causes the air to be a little diIIicult to
draw through, but while the diver is swimming on his back the valve is placed well below the level oI his Iace, and
causes a leakage oI air through his nose and dive mask. The aqualung to be described avoids the worst oI this trouble
by having separate demand and reducing valves, and the demand valve is worn in the best possible compromise
position, high on the diver's chest. The same conditions dictate that the inlet and exhaust tubes should terminate close
together at the demand valve; iI they were only a Iew inches apart there would be some positions oI the diver in the
water where the demand valve would be at a greater depth and pressure than the no-return valve in the exhaust tube,
and there would be a continuous waste oI air.

Fig. 13. - A side view of a swimmer wearing the complete aqualung equipment.
Cylinder, Closure Valve, Reducing Valve and Gauge
The Iollowing descriptions oI equipment to be bought and the modiIications required will deal in logical order with the
high-pressure gear Iirst, then the reducing valve and the low-pressure demand valve, breathing tubes, and exhaust
valve. The items to be bought and sources oI supply are given on page 166 in the "Parts List".
The high-pressure oxygen regulator, Fig. 1, should Iirst be unscrewed Irom its small cylinder, and this may prove to be
a job Ior the local garage iI the thread is obstinate. Then the complete reducing valve, with its hat-shaped cover, should
be unscrewed, revealing the microscopically small oxygen aperture in the valve outlet. This jet should be removed by
unscrewing the the aperture soldered up. Care should be taken not to lose the copper washer which seals the jet in


Fig. 1. - 1he R.A.F. oxygen regulator in its original cylinder.
1he closure valve, reducing valve and gauge are unscrewed as
one unit and fitted into a much larger cylinder, which although
capable of holding 4 cu. ft. of air weighs only 14lb.
Fig. 2. - Modification to the inlet side of the low-pressure valve.
place, Ior both jet and washer will have to be replaced each time the Iinal cylinder is sent away Ior reIill or test. This is
necessary because it would be impossible to Iill the cylinder iI there were a continuous leak through the jet seating.
However, when the complete aqualung is assembled the normally closed demand valve prevents this escape oI air.
The heavy spring against which the metal diaphragm oI the reducing valve operates is a steel spring, and thereIore
liable to rust. It should be removed and either plated with a corrosion-resistant metal, or else well greased and kept well
greased. It may be taken out aIter removing the hat-shaped cover and unscrewing the large hexagon lock-nut and
hexagon nut. Alternatively the cover may be waterprooIed by soldering a path over the 3/8 in. diameter hold and
seating the cover on to a rubber washed.
The other components which Iorm an integral part oI the high-pressure oxygen regulator are a closure screw-valve, a
gauge calibrated 0 to 10 minutes, a Iemale bayonet outlet connection and a reIilling connection which is normally
closed by a blind 3/4in. B.S.P. nipple.
The gauge reads 10 minutes when the cylinder is Iilled to 1,800lb. per sq. in., so at the maximum permitted Iilling
pressure oI 1,980lb. it will read just above this Iigure, and when the cylinder is only halI Iull it will read just above the
Iive-minute mark. It is important to have this gauge sealed up eIIiciently with a good adhesive waterprooI resin, Ior it
will have to withstand water pressure oI pounds per square inch without the slightest leak.
The Cylinder
The regulator head must now be stripped as completely as possible oI all its Iragile components, and their sealing
washers, and the brass body sent away to the suppliers oI the large cylinder, who will Iit it into the cylinder neck. This
is a specialist's job and cannot be attempted by anyone with only normal workshop Iacilities.
The cylinder, which is by Iar the most expensive item in the whole assembly, is a specially made lightweight model,
19in. tall and 7in. diameter. Although the wall thickness is only about 1/10in. the cylinder will withstand test pressures
oI 3,000lb. per sq. in., Ior it is solid drawn Irom manganese steel, strongest oI the steel alloys.
Reducing Valve-Demand Valve Connector
From the reducing valve which, is carried on the cylinder, air at about 100lb. per sq. in. pressure has to be taken via
Ilexible hose to the demand valve on the diver's chest. The rubber tube which is supplied with the double-ended
bayonet connector is not strong enough to withstand this pressure, and the high-pressure hose which is recommended
should be used. This hose is tested at 3,000lb. pressure when new, so the Iactor oI saIety should be adequate.
In order to Iit the push-on nozzle oI the male bayonet to the hose it will probably be necessary to dismantle the Iitting
and turn down the end oI the nozzle slightly. Note that the perIorated nozzle will not unscrew Irom the thread. When
the high-pressure hose has been pushed well on to its nozzle a Jubilee clip should be screwed on to ensure a thoroughly
saIe connection.
The Iemale end oI the double bayonet connector should be dismantled, and the outer tube put on one side Ior
transIormation into a no-return valve. The rounded end oI the push-on nozzle should again be turned down so that it
can be Iorced into the high-pressure hose, and its other end should be shaped and threaded 1/8in. B.S.P. as shown in
Fig. 2. The demand valve, which comes next in the sequence, consists oI a modiIied Calor gas low-pressure valve 1946
Type F, and the inlet should be tapped 3/8in. B.S.P. to take the nozzle which has just been threaded. The large hexagon
nut should be cut away as it is not required. A thin rubber washer between the nozzle and its screw-in seating completes
a good air-tight joint.
The Demand Valve
This valve should now be dismantled aIter removal oI the eight securing bolts and the small screw-on cap. The large
compression spring and its packing piece are not required, and the Iabric diaphragm should be replaced by one oI
identical size cut out oI 1/32in. sheet rubber. Fig. 3 shows the main modiIications, which consist oI drilling the centre
oI the brass diaphragm centre, tapping it 4 B.A. and securing to it loosely a double-ended soldering tag, which is
required to hold one end oI a diaphragm tensioning spring.

The spring was close-wound with a 1/4in. Iormer on a power-driven lathe, and was originally a copper-plated steel
spring Irom a motor-cycle carburetter.The wire diameter is 0.035in. and there are 8 turns. The other end oI the spring is
attached loosely to a tack seated in a central counter-sunk hole in the screw tension-adjusting disc. The pointed end oI
the tack, which protrudes through the underside oI this disc, is looped to hold the spring end. Note that the spring
should not be Iinally connected at both ends until the disc is screwed back into its thread, but aIter doing this both ends
oI the spring should be hooked so that they cannot jump oII.
The top cover oI the valve should now be drilled with about 24 1/4in. diameter holes, to allow the water to reach the
side oI the diaphragm remote Irom the air supply. The single hole that already exists is not suIIicient because there is
considerable movement on the diaphragm which will tend to displace a cubic inch or so oI water at each movement,
and any restriction on exit or ingress oI water would make the action oI the diaphragm sluggish.
The air outlet Irom the demand valve must be enlarged, but beIore this is done the lever mechanism will have to be
removed Irom the channel in the base oI the valve. It is not easy to remove the lever pivot, but this will not prove to be
an obstacle iI care is taken in enlarging the hole. The size oI the hole will vary according to the size oI outlet tube to be
used, but it should be remembered that the bore oI this breathing tube section should not at any point be less than 3/4in.
The outlet shown in Fig. 4 was made Irom a piece oI aluminium alloy tube Iound in the scrap box, and as the outer
diameter oI the thicker section is 1in. it connects well to the respirator tubing which take the air to the mouth. The
narrow section was threaded 3/4in., 26 t.p.i., and the body oI the valve was tapped accordingly. BeIore being Iinally
screwed home the threads were smeared with waterprooI resin varnish.


Fig. 3. - Arrangement for tensioning the diaphragm to keep the
low pressure inlet valve normally closed.

Fig. 4. - Connection for
the inlet breathing tube.

Fig. 5. - Photograph of the completely
modified low-pressure valve.







It may prove possible to use a short length oI the 3/4in. copper water pipe which is recommended later Ior the
mouthpiece; iI so a small rubber adaptor tube oI the type described then will ensure a good Iit Ior the respirator tube. A
circular clip held by one oI the eight bolts oI the demand valve cover is used to secure a short piece oI the same 3/4in.
copper tube, on to which the outlet end oI the exhaust tube is pushed. Iig. 5 is a photograph oI the completely modiIied
demand valve.
The valve illustrated was mounted on a 4 1/2in. by 3 1/2in. piece oI 1/16in. thick Iibre sheet, and two edges oI the Iibre
were slotted so that small webbing straps could be secured to it. It was just possible to drill the rear oI the valve, at the
Iour corners oI the channel, and tap the holes 4 B.A. to screw the Iibre to the valve, without risk oI the holes breaking
through into the air chamber.
Mouthpiece
The inlet respirator tube connects between the outlet oI the demand valve and the mouthpiece, which as has already
been stated is made Irom 3/4in. copper pipe. Rubber adaptor sleeving with inside and outside diameters oI 3/4in. and
1in. respectively is used to ensure a tight non-slip Iit. In the aqualung shown in the photographs it was decided to bring
the high-pressure hose over the right shoulder, although it could equally well have gone over the leIt, and this
positioned the inlet air tube on the leIt oI the wearer, with the exhaust tube on the right. For this reason the leIt-hand
side oI the mouthpiece is the longer side. This has been arranged so that both inlet and exhaust tubes will be at similar
tensions Ior any position oI the head, and compensates Ior the Iact that the securing Iixing Ior the exhaust tube on the
demand valve is slightly above that Ior the inlet tube.
it must be remembered that the normal position oI the head oI a submarine swimmer is tilted back as Iar as it will
comIortably go, so that the eyes can look along the line oI travel. This dictates the odd-looking angle oI the T-piece
pipe which is soldered on to the main body oI the mouthpiece; when the eyes are looking straight up, the two ends oI
the pipe should be almost parallel with the chest. Perhaps the importance oI this will be lost to readers with no
experience oI diving masks; those who have used them will know that all models available at present have a very
limited Iield oI view. Fig. 6 is a sketch oI the mouthpiece tubing.
A second advantage possessed by this aqualung over conventional models may be mentioned here. When the head is
thrown back in this way to give good visibility in the direction oI travel a demand valve carried on the cylinder head
proves to be most uncomIortable to the back oI the head; the demand valve worn on the chest solves this problem

Fig. . - Shaping of 3/4 in. copper tube to form the
mouthpiece. 1he rubber "gag" which fits onto the
short arm can be taken from a conventional snorkel
tube.


Fig. 7. - Details of a simple and effective no-return valve for the air
exhaust tube.


completely.
One oI the cheap snorkel tubes readily available in sports shops will supply the rubber mouth-grip which puts the
Iinishing touch to the mouthpiece, and incidentally it will be a good thing always to take this snorkel along when diving
in open water. The aqualung with an exhausted air cylinder soon leads to an aqualung with an exhausted swimmer, Ior
the weight oI the lung plus the lead weights worn on the belt keep a surIace swimmer very low in the water.
No-return Valve
A no-return valve must be Iitted in the exhaust tube, close to its termination on the demand valve, to prevent water
being drawn up the tube. The type Iound in the top oI a Services respirator charcoal Iilter is Iairly satisIactory, but
many minutes' use may allow a noticeable amount oI water to get into the tube.
The design shown in Fig. 7 can be quite easily made, mainly Irom the outer shell oI the Iemale bayonet connector
which has already supplied the high-pressure inlet nozzle Ior the demand valve. The slotted end oI the tube is cut oII,
and both raised shoulders are turned down, leaving a tube with a smooth plain exterior and a slight constriction at one
end oI the interior. A piece oI the thin rubber sheet which was used Ior the demand valve diaphragm must be cut so that
it rests inside the tube on this constriction and yet does not quite touch the inner wall oI the tube at any point. A U-
shaped piece oI phosphor-bronze spring wire is next pushed down the tube to hold the rubber in place, and this is held
Iirmly, in turn, by two square pillars cut Irom 3/16in. thick rubber. The pillars are nicked along their length to receive
the vertical arms oI the U-spring. A circular Iibre or rubber washer, which must be a tight push-Iit in the tube, Iinally
secures the whole assembly. ReIerence to the accompanying drawing will make the build-up clear.
Webbing Harness
The aqualung is carried on the back oI the diver, and its 20lb. or so oI weight in air must be supported Iirmly by sturdy
shoulder straps. In water, however, the cylinder is buoyant, and its movement away Irom the body in any direction
must be prevented. In addition it makes the diver buoyant, and he must carry anything Irom 8lb. to 14lb. oI lead
weights (depending on his own buoyancy and the density oI the water) on the Iront oI his belt to give him neutral
buoyancy. This is important, Ior aqualung divers should be able to swim in all directions in the water with equal ease.
Note that as water is virtually incompressible its density does not vary appreciably Irom surIace to bottom,
notwithstanding the well-worn tale about drowned people's bodies Iloating about eternally, each at its own level. In
actual Iact not even a skeleton has ever been Iound in wrecks which were more than a Iew months old.
The harness belt must be strong in order to carry the lead weights, and it must be Iirmly buckled so that it does not Ily
open in the water. On the other hand there may come a time when the whole aqualung proves and encumbrance and has
to be shed quickly in the water, so the buckle must be capable oI quick release. II the Services type oI belt suggested in
the parts list is used, it will be Iound possible to hammer the male end oI the buckle until it grips securely and get gives
this quick release Iacility. Time spent in practising opening and closing the buckle by touch many times beIore entering
the water will be time well spent.
Fig. 8 shows the conIiguration oI the webbing harness. When ordering the 3/4in. straps Ior the weights and cylinder
top, speciIy brass buckles, Ior these are used in the harness, and must be corrosion resistant. The buckles on the 1in.
straps are zinc-plated, and so are Iairly corrosion-resistant, but they do not give such smooth and easy adjustment at
those places where the harness has to be Iitted tightly where it is already reasonably taut as the recommended "D"
rings. These are 1in. D-shaped rings, sewn into strap ends in twos to make a continuously-adjustable buckle. The
method oI use is to pass the Iree end oI the strap which is to be joined down through both rings and then down again
between them. A tug on the Iree end oI the strap will then be Iound to tighten the Iastening, but there will be no slip
back in the other direction.
All joins in the webbing straps and belt should be double-stitched (using two needles) with carpet thread, and the Iree
ends oI the thread tied in a Iirm knot at the end oI each rectangular stitching section. Cut ends oI strap should be tucked
under Ior about 1/2in. beIore stitching is begun, in order to prevent unravelling oI the end threads. Unravelling oI the
small straps used Ior the weights can be prevented iI a waterprooI glue is rubbed well into each side oI the place it is
intended to cut, an hour or two beIore cutting them to size. One oI these spare cut ends is used Ior halI oI the cylinder-
securing strap near the cylinder head, and two more hold the demand valve high up on the diver's chest. This chest strap
must also be quick-release, opening at the side opposite the side Irom which the high-pressure hose comes. The strap
illustrated was Iitted with a sturdy press-stud Iastener taken Irom a Services respirator haversack.


The Lead Weights
Six oI these were cast, each between 2 and 2 3/4lb. in weight. Scrap lead was used, and aIter being scrubbed it was
weighed out into batches ready Ior melting. The melt was kept clear oI scum, by skimming and stirring until no more
scum was produced, and then poured into the mould shown in Fig. 9. The Iour walls oI the mould were made by
tacking together Iour pieces oI 2 1/4in.x 3/4in. wood, two oI which had metal strips screwed in vertically and centrally
to Iorm the strap channels. AIter the wood had been dried thoroughly in Iront oI a hot Iire Ior some hours it was placed
on a trowel, which Iormed the base oI the mould. Fig. 10 is a photograph oI a complete set oI weights.

Testing
The tension on the demand valve adjusting spring should be adjusted until it is impossible to blow through the inlet
tube to the valve, although the valve has just been opened by sucking air through the outlet tube. AIter assembling the
rest oI the aqualung, it should be ready Ior Iull-pressure testing.
The cylinder should be delivered to a British Oxygen Company air-pumping station, iI at all possible, to avoid possible
risk oI damage to the valve-gear by larger cylinders during transportation Irom collection depot to pumping station. In
any case it is wise to Iit some kind oI protective cover to the valve-gear at the top oI the cylinder. The bayonet

Fig. 12. - Rear view of the aqualung.

Fig. 8. - Details of the harness, assembled from khaki webbing straps,
and belt.

Fig. 9. - Suggested mould shape for casting the lead weights.

Fig. 1. - A set of lead weights totalling 13 1/2lb
connector at the end oI the high-pressure hose must, oI course, be disconnected, and the soldered-up jet replaced in its
position in the reducing valve.
For a Iew shillings extra charge B.O.C. will paint the cylinder in the regulation Iashion, with alternate black and white
quadrants at the top third and grey below, with the word "AIR" stencilled on plainly in black. The cost oI an air Iill is
about 4s. Ior commercial quality, but more than twice this Ior medically pure air.
II you should be able to join a branch oI the British Sub-aqua Club which possesses a transIer pump they will be able to
pump your cylinder Ior you, at a more reasonable cost, Irom a large B.O.C. cylinder.
With the cylinder Iully charged, make sure that the closure screw-valve is turned oII (Iully clockwise) and then
unscrew the reducing valve to remove the blind jet. AIter replacing the valve and tightening well home, the bayonet
connector should be replaced, and the closure valve opened slightly and slowly.
It must be stressed here that sudden build-ups oI pressure should be avoided whenever compressed gasses are being
used. Temperatures inside the equipment will be produced by careless handling which could easily touch oII a diesel
explosion iI any traces oI oil or grease are present. Equally, oI course, it is important to keep all apparatus Iree oI even
minute smears oI oil or grease.
With the valve open there will probably be a slight escape oI air through the demand valve; this should be stopped by
increasing the tension on the spring with the adjusting screw-disc. The needle oI the pressure gauge should be a little
over the 10-minute mark, and when the closure valve is shut the needle should stay where it is. iI it does not there must
be an air leak somewhere. iI the test is satisIactory the valve may be opened again and the operation oI the aqualung
tested by taking a Iew breaths Irom the mouthpiece.
II air comes through Ireely as required and can be exhaled without eIIort the equipment is ready Ior a Iull-scale water
test, and you are well on the way to partaking oI the joys oI this new and exhilarating sport oI submarine exploration.
Figs. 11 and 12 give some idea oI what you will look like to the Iish.
First Published in: Newnes Practical Mechanics January 1955


A suitable weight belt could be easily made Irom a 12 ga. shotgun ammo belt. To make the lead weights use a 3/4 inch
water pipe (steel not copper) as a mold. Check your local junk yard or grease monkey Ior used balancing weights. To
melt them use any steel pot and just about any heat source will do. Wear leather gloves saIety goggles, and be careIul
pouring it, hot lead and water don't mix, they will react violently. I will show how to make the shotgun belt on the next
update, plus more on building a diIIerent type oI an aqualung. With a little enginuity these inexpensive on demand
regulators can be adapted to a snuba setup.




Remember! These apparatus are to be used for shallow water diving only!



This is a typical homemade aqualung . It uses two
Lancaster aircraIt oxygen tanks.